Questions and musings on EQ (and gain)

[Grant Bunter]

I'm getting back to my "How to set up a DCX2496" thread and I'm about to launch in to the EQ section.

I've been doing a lot of reading about EQ lately, and it's raised a lot of questions that I'd like some answers and/or consensus on before proceeding in that thread.

It's not just EQ however.
I recently read an article that showed me how increasing gain (and that's what EQ does if you add rather than cut) affects crossover point.

IIRC it's Radian in particular that's been saying this for ages, but this article showed how increasing gain across the board in the the sub bandpass upwardly raises crossover point.
You can use this to advantage.
Raising your gain by say 6dB, in the sub bandpass, which is not uncommon in modern systems (rather than boosting EQ 6dB) would raise your crossover point of, say 100Hz, to say 115Hz. So you can still raise gain, but lower the crossover point to 70Hz for a resultant crossover point of 100Hz (and utilise the rising response of Bill's sub designs).
This minimises EQing in that bandpass.
Comments?

I've also read extensively that using EQ alters phase response.
And as such, you should use EQ before crossover.
So in the case of DCX setup, this suggests using EQ at the "input" section rather than the individual output sections (though using EQ at the output stages would increase the amount of overall PEQ).
Comments.

Any other significant comments on either EQ or gain would be great as well, just trying to get it all in order...

TIA

[MarcBass]

G'day Grant,

Sounds like you have come accross the old technique of EQing with the crossover. Rane discuss (pay lip-service to) this in the manuals for their old RTA units. (Forget the model numbers now but can get them for you if you want.)

I have used this style for quite a while with various systems and it does work.

1. Essentially, set the crossovers as per speaker / cab spec then run a signal (appropriate freq sine wave) through each section (bass, mid, top etc.) measuring spl response of each band and adjusting gain till they "match" or give the spl required of the band if you are not aiming for a flat response.

2. Next adjust the crossover points until you get the response you want as indicated on the RTA. This is where you can figure out wheather to have a "gap" between bass and mid and what is the best filter (LR or BW) to use for the given cabs if you can change filters.

3. After the above two steps, adjust EQ to fine tune the rig to match the desired EQ "curve".

4. Run a few test CD's to check the end result with human ears. (I usually use Stealy Dan, Coors - Toss the Feather and something close to the style the band will play.) While all my work is live bands I have found lately that audiences expect DJ type service while the band is taking breaks ..... grrrrrrr!!

These days I use DBX driveracks (PA+ for FOH and the cheaper unit PX? for foldback). I have adapted the old technique to the DriveRacks with no drama. My BFM rig is incomplete with only T39's and W8's at the moment. The above technique has worked fine with T39's combined with Peavey SP2G's or Peavey HiSys (12 inch top). Have also succesfully run a single T39 with 2 HiSys units. Hence, I believe the technique is still valid with Bill's designs.

Having said that, I also believe there are many ways to skin a cat and there are lots of shades of grey, so would never sugest this is the only way. I would also point out that the only variation of highpass frequency I ever entertain on subs is to raise it. eg. if the content running through my rig does not contain, or require, big bass guitar and huge kick drum (low frequency fundamentals) I have been known to sneak the HPF up to say 50 or 60 Hz. (If I feel the need to go much higher I tend to leave the subs at home.)

Cheers

Marc

[Bill Fitzmaurice]

Raising your gain by say 6dB, in the sub bandpass, which is not uncommon in modern systems (rather than boosting EQ 6dB) would raise your crossover point of, say 100Hz, to say 115Hz.

Not exactly. Nominally you have the subs and mains both producing a -3dB signal at the crossover frequency, which then sum to 0dB. If you boost the sub level by 6dB then the sub will be +3dB at the crossover frequency, and when summed with the mains output there will be a response bump at the crossover. If anything what running the subs hot allows you to do is to raise the crossover frequency of the mains by 10-15Hz or so, reducing the load on them while eliminating that response bump. To be clear, you would probably set the low pass at 100Hz, the high pass at 110-115Hz. If both the sub and main were running at the same level that would give a response dip in the 110-115Hz region, but with the sub running hot it doesn't.

[Grant Bunter]

Cheers marcbass and Bill,
I will spend some time digesting both your comments.

I should have included the link to the article I mentioned.
http://www.prosoundweb.com/article/tuni ... _to_speed/

The example is on page one...

[Bill Fitzmaurice]

I should have included the link to the article I mentioned.
http://www.prosoundweb.com/article/tuni ... _to_speed/
The example is on page one...

The example is flawed, as the charts are inaccurate. They show the response of each speaker separately, not the summed response. Have a look through this, where you'll find charts that show not just the individual responses, but also the combined response with various filter configurations.
http://www.bcae1.com/xoorder.htm

[sine143]

Raising your gain by say 6dB, in the sub bandpass, which is not uncommon in modern systems (rather than boosting EQ 6dB) would raise your crossover point of, say 100Hz, to say 115Hz.

Not exactly. Nominally you have the subs and mains both producing a -3dB signal at the crossover frequency, which then sum to 0dB. If you boost the sub level by 6dB then the sub will be +3dB at the crossover frequency, and when summed with the mains output there will be a response bump at the crossover. If anything what running the subs hot allows you to do is to raise the crossover frequency of the mains by 10-15Hz or so, reducing the load on them while eliminating that response bump. To be clear, you would probably set the low pass at 100Hz, the high pass at 110-115Hz. If both the sub and main were running at the same level that would give a response dip in the 110-115Hz region, but with the sub running hot it doesn't.

bill, dont you mean you want both subs and mains producing a -6db signal at the crossover frequency to sum to 0db? a -3db signal for both would sum to +3dB.

[Bill Fitzmaurice]

bill, dont you mean you want both subs and mains producing a -6db signal at the crossover frequency to sum to 0db? a -3db signal for both would sum to +3dB.

Yes, my bad. -6dB for both sums to zero; that's a Linkwitz-Reilly alignment. -3dB for both is Butterworth, which sums to +3dB. We use Chebychev filters in our horn loaded cabs, which technically sums to +6dB, but you don't see it as a +6dB summing, because the response of the LF section is rolling off in the vicinity of the crossover. If we used a lower Q filter you'd see more of a response dip in the 1.5kHz-2kHz range.

[sine143]

on the topic, whats the deal with Bessel filters. many of the venues I've walked into that will actually let me peer into their dsp settings seem to be using them (and none of the techs can explain why).

[Bill Fitzmaurice]

on the topic, whats the deal with Bessel filters. many of the venues I've walked into that will actually let me peer into their dsp settings seem to be using them (and none of the techs can explain why).

It's just a different Q, -5dB at fc, summing to about +1dB. If the techs don't know why they're using them chances are that was the default alignment.

[Grant Bunter]

Thanks sine and Bill, I was going to ask if BW or LR made a difference to summing.
Now answered

[Radian]

IIRC it's Radian in particular that's been saying this for ages, but this article showed how increasing gain across the board in the the sub bandpass upwardly raises crossover point.
You can use this to advantage.
Raising your gain by say 6dB, in the sub bandpass, which is not uncommon in modern systems (rather than boosting EQ 6dB) would raise your crossover point of, say 100Hz, to say 115Hz. So you can still raise gain, but lower the crossover point to 70Hz for a resultant crossover point of 100Hz (and utilise the rising response of Bill's sub designs).
This minimises EQing in that bandpass.
Comments?

To quote Siegfried Linkwitz:

"A very important point that people sometimes miss in this discussion is that when we are speaking of a given crossover, we are talking about an acoustic crossover, or what happens acoustically. Now, what I have to do electrically to achieve the correct acoustic response may not look anything at all like a textbook filter design. The actual filter often looks very little like the drawings I may show to explain any given example. This is also true for a Butterworth filter. It is highly unlikely that a textbook electrical Butterworth crossover will produce an acoustic Butterworth response, because the driver's response enters into the picture as well."